Biomacromolecules 2005, 6, 2776-2784
2776
Force Measurement for Antigen-Antibody Interaction by Atomic Force Microscopy Using a Photograft-Polymer Spacer Alimjan Idiris,† Satoru Kidoaki,‡ Kengo Usui,† Tei Maki,† Harukazu Suzuki,§ Masayoshi Ito,§ Makoto Aoki,† Yoshihide Hayashizaki,§ and Takehisa Matsuda*,‡ Core Research for Evolutional Science and Technology, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Division of Biomedical Engineering, Graduate School of Medicine, Kyushu University, Fukuoka 812-8582, and Laboratory for Genome Exploration Research Group, RIKEN Genomic Sciences Center, RIKEN Yokohama Institute, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, Kanagawa 230-0045, Japan Received April 13, 2005; Revised Manuscript Received June 14, 2005
To determine the intermolecular force on protein-protein interaction (PPI) by atomic force microscopy (AFM), a photograft-polymer spacer for protein molecules on both surfaces of the substrate and AFM probe tip was developed, and its effectiveness was assessed in a PPI model of a pair of human serum albumin (HSA) and its monoclonal antibody (anti-HSA). A carboxylated photoiniferter, N-(dithiocarboxy)sarcosine, was derivatized on both surfaces of the glass substrate and AFM probe tip, and subsequently water-soluble nonionic vinyl monomers, N,N-dimethylacrylamide (DMAAm), were graft-polymerized on them upon ultraviolet light irradiation. DMAAm-photograft-polymerized spacers with carboxyl groups at the growing chain end but with different chain lengths on both surfaces were prepared. The proteins were covalently bound to the carboxyl terminus of the photograft-polymer chain using a water-soluble condensation agent. The effects of the graft-spacer length on the profile of the force-distance curves and on the unbinding characteristics (unbinding force and unbinding distance) were examined in comparison with those in the case of the commercially available poly(ethylene glycol) (PEG) spacer. The frequency of the nonspecific adhesion force profile was markedly decreased with the use of the photograft spacers. Among the force curves detected, a high frequency of single-peak curves indicating the unbinding process of a single pair of proteins and a very low frequency of multiple-peak profiles were observed for the photograft spacers, regardless of the graft chain length, whereas a high frequency of no-force peaks was noted. These observations were in marked contrast with those for the PEG spacer. The force peak values determined ranged from 88 to 94 pN, irrespective of the type of spacer, while the standard deviation of force distribution observed for the photograft spacer was lower than that for the PEG spacer, indicating that the photograft spacers provide a higher accuracy of force determination. Introduction Protein-protein interaction (PPI) plays various critical roles in a wide range of biological events at intracellular, intercellular, and tissue levels. The mechanistic understanding of PPI is a primary subject in the bioscience field because PPI triggers major functions of biological activities. Besides information on kinetic and thermodynamic parameters such as binding and dissociation rate constants and the equilibrium dissociation constant, the direct determination of the force on the mechanical dissociation of an associated pair of two different proteins, termed the “unbinding force”, is an effective strategy of characterizing the nature of the intermolecular interaction. Atomic force microscopy (AFM) makes it possible to directly measure the intermolecular forces between biomolecules of interest at the single molecular level in aqueous * To whom correspondence should be addressed. E-mail: matsuda@ med.kyushu-u.ac.jp. † Japan Science and Technology Agency. ‡ Kyushu University. § RIKEN Yokohama Institute.
media down to the few piconewtons (pN) range. Many research groups have applied this method in detecting specific molecular recognition events with very weak to relatively strong interactions among biological materials, such as between the ligand and receptor, biotin and avidin, and antibody and antigen pairs.1-9 To determine such an intermolecular force, one species of protein molecules is immobilized on an AFM probe tip and the other species on the substrate. The following technical requirements must be satisfied to immobilize proteins on these surfaces: (a) avoidance of surface-induced denaturation and mechanicalforce (or indentation)-induced conformational deformation, (b) minimized nonspecific adhesive interactions with the bare probe surface or bare substrate surface, and (c) enhanced motional freedom of proteins after their immobilization on the surfaces to increase the binding chance. Spacer chemistry has been proposed to meet these requirements, which includes end-functionalized, flexible polymeric spacers such as poly(ethylene glycol) (PEG)10,11 and carboxymethylamylose.12-15 Although the significance of spacers in AFM force-distance (f-d) curve measurement
10.1021/bm0502617 CCC: $30.25 © 2005 American Chemical Society Published on Web 07/22/2005
Intermolecular Force Measurement for PPI by AFM
has been noted, there is still room for improvement in the design of spacers. In this study, the surface photograft architecture enabling the binding of a target protein to the end of the graft chain on both the substrate and AFM probe tip was explored by iniferter-based photograft polymerization.16-18 Since an iniferter, which is a molecule with a triple role as an initiator, a transfer agent, and a terminator in the polymerization process, can induce quasi-living radical polymerization only during photoirradiation, the length of the polymerized chains or the thickness of the photograft layer can be controlled by reaction conditions such as UV light intensity or irradiation time.19,20 We chose the carboxylated iniferter N-(dithiocarboxy)sarcosine (DTCS), and derivatized it on both the AFM probe tip and glass substrate, from which the water-soluble flexible polymer poly(dimethylacrylamide) (PDMAAm) with a carboxyl group at the growing chain end was graftpolymerized. The protein immobilization strategy employed links an antibody to a substrate surface between amino groups present in the lysine residues of the protein molecule and the carboxylated terminus of the photograft-polymerized chain using a water-soluble condensation agent. Because the Fc region of the antibody molecule is relatively lysine-rich, there may be some preferential orientation of the antibody, so that the availability of the antigen-binding site is enhanced.21 Using this controlled photograft spacer, intermolecular force measurements were performed between human serum albumin (HSA) and its monoclonal antibody (anti-HSA) as the pair in the PPI model. The effects of the graft-spacer length on the profile of f-d curves and the distributions of unbinding force values are discussed in comparison with those determined using the commercially available PEG spacer. The advantageous features in using photograft spacers include the avoidance of nonspecific adhesion, a clear-cut recognition of the single-pair force of antigen-antibody, the differentiation of multiple peaks in a single f-d curve, and improvement of the accuracy of the measured force value. Further improvement of the intermolecular force measurement using photograft-spacer chemistry is also discussed. Experimental Section Materials. The following commercially available materials of special reagent grade were used: HSA (Sigma Chemical Co., St. Louis, MO), monoclonal mouse anti-HSA (mouse isotype 1C8-IgG1; HyTest, Ltd., Turku, Finland), N-hydroxylsuccinimide (NHS; Wako Pure Chemical Industries, Ltd., Osaka, Japan), (chloromethylphenyl)ethyltrichlorosilane (Shin-Etsu Chemical Co., Ltd., Tokyo, Japan), (3-aminopropyl)triethoxysilane (Shin-Etsu), (3-mercaptopropyl)triethoxysilane (Shin-Etsu), poly(ethylene glycol)-R-maleimide ω-NHS ester (NHS-PEG-MAL; MW 3400; Nektar Therapeutics, Huntsville, AL), DTCS (Dojindo Laboratory, Kumamoto, Japan), DMAAm (Wako), and 1-ethyl-3-(3dimethylaminopropyl)carbodiimide hydrochloride (EDC; Sigma). Solvents and other substances, all of which are of special reagent grade, were used after appropriate purification. Water was deionized using a Milli-Q reagent water
Biomacromolecules, Vol. 6, No. 5, 2005 2777
system (Nippon Millipore, Ltd., Tokyo, Japan) to 18 MΩ cm resistivity (DI water). DTCS Derivatization of Glass Substrates and AFM Probes. Both glass substrates (0.12-0.17 mm in thickness, 15 mm in diameter; Matsunami Glass Industries, Ltd., Osaka, Japan) and AFM probe tips (a Si3N4 triangular cantilever with a sharpened pyramidal tip; OMCL-TR400PSA-1, Olympus Optical Co., Ltd., Tokyo, Japan) were treated with DTCS according to the method previously reported by us19 (see Scheme 1). Briefly, for the derivatization of glass substrates, the following were carried out: (1) sonication in a 0.5% aqueous solution of Triton X-100 and DI water, (2) immersion in 80 °C piranha solution (concentrated H2SO4: 30% H2O2 ) 7:3) for 1 h, (3) sequential rinsing with DI water, acetone, a 1:1 solution of acetone and toluene, and toluene, (4) immersion in a 5% (v/v) toluene solution of (chloromethylphenyl)ethyltrichlorosilane for 18 h with shaking under an argon atmosphere at room temperature and rinsing with acetone and toluene (these silanization and rinsing steps were repeated two times), (5) after chloromethylation treatment, sequential rinsing with toluene, acetone, and DI water, and drying at 115 °C for 10 min in air, (6) immersion of the treated glass substrates in a 5% methanolic solution of DTCS for 18 h at room temperature, thorough rinsing with methanol and DI water, air-drying, and storage in a dark desiccator. For the derivatization of AFM probe tips, procedures 4-6 were performed with great care without any shaking throughout all the steps to avoid any mechanical damage to the cantilevers. Surface Photograft Polymerization. The photograft polymerization of DMAAm on the DTCS-modified glass surface and AFM probe tips was performed according to our method previously reported.19 Briefly, after a methanolic solution of DMAAm (1.0 M) was bubbled with nitrogen, 300 µL of the monomer solution was dropped onto the DTCS-glass substrate placed at the center of the Teflon slice chamber with a 25 mm diameter hole (0.6 mm in thickness and 28 mm in outer diameter) in the glass vessel (35 mm in diameter). DTCS-modified AFM probe tips were placed in the Teflon slice chamber with small rectangular wells (3 mm × 7 mm). Then the chambers were covered with sapphire glass (25 mm diameter) under a nitrogen atmosphere (∼100 µm solution thickness). UV light (200 W Hg-Xe lamp, Spot cure model SP-V, Ushio Inc., Japan) was irradiated at room temperature (0.96 W/cm2 intensity measured at a wavelength of 325 nm). After the irradiation, the glass substrates and AFM probe tips were repeatedly rinsed with ethanol and DI water and then dried under nitrogen gas. Protein Immobilization with Photograft-Polymer Spacers. The covalent immobilization of HSA and anti-HSA was performed as follows: (1) Both the photograft-polymerized glass substrates and AFM probe tips were immersed in a 1:9 mixture solution of EDC (25 mg/mL in 0.5 M NaH2PO4, pH 4.5) and NHS (2.0 mg/mL in 1,4-dioxane) for 15 min at room temperature to form the activated ester group at the terminus of the graft chain. (2) After thorough rinsing with DI, the glass substrates were immersed in phosphatebuffered saline (PBS; pH 7.4) containing anti-HSA (1.0 mg/
2778
Biomacromolecules, Vol. 6, No. 5, 2005
Idiris et al.
Scheme 1. Schematic Representation of the Experimental Procedure for Derivatization of a Glass Substrate with DTCS and Photograft Polymerization of DMAAm on a DTCS-Modified Surface
mL) and the AFM tips were immersed in PBS containing HSA (1.0 mg/mL) for 1 h at room temperature. (3) After thorough rinsing with PBS, these protein-fixed glass substrates and AFM probes were immersed in 1.0 M Tris-HCl buffer (pH 7.5) for 15 min at room temperature to hydrolyze the activated ester group. Protein Immobilization with a PEG Spacer. HSA and anti-HSA were fixed on both the glass substrate and AFM probe surfaces modified with the heterobifunctional PEG spacer with the activated ester group at one end and maleimide at the other end (NHS-PEG-MAL) according to a method previously reported:22 (1) The glass substrates and AFM probes were treated with (3-mercaptopropyl)triethoxysilane with the same procedure of silanization in DTCS modification described above. (2) The sulfhydrylated glass substrates and AFM probes were treated in the 20 mM ethanolic solution of NHS-PEG-MAL for 1 h at room temperature. (3) The protein immobilization procedures and conditions were the same as those for the graft spacer. Force-Distance Curve Measurements. The f-d curves were measured using the Molecular Force Probe 3D (MFP3D; Asylum Research, Santa Barbara, CA). Silicon nitride cantilevers with 200 µm in length and with a nominal spring constant of 0.02 N/m were used. Actual spring constants for all the cantilevers used were determined by the thermal noise/ resonance method.23 The measurements were carried out in PBS at a constant piezo scan velocity of 500 nm/s. In each force measurement, around 4000-5000 approaching/retracting f-d curves were collected using several independently prepared protein-immobilized probes, and all of the observed f-d curves were recorded continuously over 200-300 different lateral positions on the surface of each sample. The apparent graft thickness of the photograft-polymerized layer, the repulsive interaction distance in the tip-approaching
traces of f-d curves, was measured from the traveling distances of the tip between the initial contact position of the tip/graft layer and the fully compressed position of the graft layer. The former was defined as where the force increment in the tip-approaching process becomes significantly higher than the noise fluctuation level in the tipsample noncontact region, and the latter as where AFM tip movement is stopped due to the hard surface of the fully compressed graft layer. The chain length of the photograft spacer was determined from the tip-retracting traces of the f-d curves in the polymer bridged between the AFM probe tip and glass substrate according to the following procedure: (1) the AFM probe tips were treated with (3aminopropyl)triethoxysilane, and (2) the carboxyl groups of the DTCS at the termini of the photograft-polymerized DMAAm chains on glass substrates were bound to react with amino groups derivatized on the AFM probe tip in the presence of EDC and NHS. Results and Discussion Preparation of a Photograft-Polymer Spacer with a Carboxyl Group at the End. According to our previous method,19 a water-soluble photograft polymer (PDMAAm) with the carboxyl terminus on both the AFM probe tip and glass substrate was sequentially prepared by photoiniferter polymerization techniques as shown in Scheme 1 (the detailed analyses of each step of reactions were reported in our previous paper19). Under appropriate reaction conditions, photolysis, recombination, and monomer addition reaction occur under UV light irradiation. In such a situation, an iniferter group always exists at the growing chain end and the graft-chain length increases with the photoirradiation period. Therefore, photopolymerization behaves similar to “quasi-living” radical polymerization.
Intermolecular Force Measurement for PPI by AFM
Biomacromolecules, Vol. 6, No. 5, 2005 2779
Figure 1. Representative approaching traces of f-d curves between a bare AFM probe tip and a series of DMAAm-graft-polymerized glass surfaces (a) and between a bare glass substrate and a series of DMAAm-graft-polymerized AFM probes (b). t indicates the photoirradiation period (min). The zero position of the probe tip is defined as the starting position of the linear part in the tip-sample contact region.
To determine whether the thickness of the photograftpolymerized layer of DMAAm increases with photoirradiation time, f-d curve measurements were carried out for the glass substrates and AFM probes, both of which were photopolymerized for different photoirradiation periods. Two series of f-d curves, between a nontreated AFM probe tip and photopolymerized glass substrates (Figure 1a) and between a nontreated glass substrate and photopolymerized AFM probes (Figure 1b), were obtained in PBS. In the approaching trace of the f-d curves, repulsive forces due to steric interaction derived from a water-swollen graft layer were observed for both cases in which the repulsion behavior largely depended on the photoirradiation period. Irrespective of the case, the repulsive interaction distance, which reflects the thickness of the water-swollen graft layer,20 gradually increased with the photoirradiation period, indicating the photoirradiation-period-dependent growth of the photograftpolymer chains. This behavior coincides with that observed in our previous study: a linear increase in graft thickness with photoirradiation period.19 On the other hand, a much larger repulsive interaction distance was observed on the glass surface at a given photoirradiation period than that on the AFM probe surface (Figure 1b). For example, to fabricate a photograft layer with an apparent thickness of 85 nm, it took photoirradiation periods of 3 min for the glass substrate and 7 min for the AFM probe. This difference may be attributable to the difference in the surface-derivatized density of DTCS between the glass substrate and Si3N4 AFM probe tip. In the derivatization of DTCS, (chloromethylphenyl)ethyltrichlorosilane was precoupled to SiOH groups on both of the substrates which were introduced through the piranha treatment. The degree of introduction of SiOH groups should be rather low on the Si3N4 surface of the AFM probe tip compared with the SiO2 surface of glass substrates. These
Figure 2. Representative f-d curve between DMAAm-graft-polymerized AFM probes and glass substrates. t indicates the photoirradiation period (min).
differences in surface chemical conditions may result in the difference in the surface derivatization degree of DTCS. As shown in Figure 2, the f-d curve measured between the DMAAm photograft-polymerized AFM probes (photoirradiation period t ) 6 min) and photograft-polymerized glass substrates (t ) 3 min) showed that almost no interlayer adhesion force between the PDMAAm photograft layers was observed (observed frequency
